‘Statistical consulting’ in practice
In the first semester of the 2022/23 academic year, the course ‘Statistical Consulting’ was completed on the program ‘Applied Statistics with Network Analysis’. Its main goal was to immerse students in real business problems and familiarize them with the rules and principles of a consultant's work in the industry.
This academic year, the partner of the program is Zoom TV, which develops applications for SmartTV. Its founder and CEO, Mikhail Baklanov, challenged students to identify the relationship between the quality of a recommendation system and the business metrics of an application for watching TV channels.
As in any real project, the work of the students began with a briefing with the customer, after which they split into project teams and started looking for solutions.
Each team had to analyze the provided data, prepare and submit an analytical report consisting of 5 main blocks:
- Formulation of the problem
- Goals and objectives of the project
- Project methodology
- Descriptive statistics
- Model description
- Results
- Team
Students had to use the knowledge gained in the courses of the programme ‘Applied Statistics with Network Analysis’ (‘Applied Linear Models’, ‘Time Series Analysis’, ‘Categorical Variable Analysis’, ‘Digital Data Analysis’, ‘Machine Learning’, etc.) and use R or other data manipulation software. Despite the fact that all teams received the same task, they approached the solution of the task in different ways, based on correlation, regression, discriminant and other types of analysis.
Mikhail Baklanov noted that the students did an excellent job. According to him, "Some groups were able to surprise in terms of the approach taken and the depth of the results obtained, given that the ZoomTV analysts were also working on solving the problem."
